This study aims to select soybean lines with desirable nutraceutical characteristics under specific meteorological conditions in the selection environments. The experiment was conducted during the 2021/2022 harvest at Escola Fazenda, Regional University of the Northwest of the State of Rio Grande do Sul, located in Augusto Pestana- RS, Brazil. The experimental design was augmented blocks, with 27 populations of F2 generation, 118 lines of F5 generation, 70 lines of F8 generation, and 15 soybean cultivars as interim checks, arranged in three replications. The chemical composition of soybeans was evaluated using Near Infrared Spectroscopy - NIRS method to analyze various nutritional components. The IRC050 genotype exhibited a nutraceutical profile characterized by an increment in stearic fatty acid (STEAR) and reduction in fiber (FB) and mineral material (MM) contents. Among the F5 generation, the following genotypes were selected based on their nutraceutical traits: L174F5, L307F5, L322F5, L205F5, L190F5, L196F5, L219F5, L183F5, L301F5, L172F5, L198F5, L182F5, L319F5, L21F5, L188F5, L217F5, L302F5, L315F5, L36F5, L318F5, L211F5, L165F5, L321F5, and L179F5. The indirect selection of genotypes with higher levels of linoleic acid and protein was facilitated using the enzyme peroxidase.